A custom mRNA vaccine, made from one patient's tumor, delayed melanoma's return in a late-stage trial

A custom-built mRNA vaccine, designed from a sample of one patient's own tumor and given alongside Merck's Keytruda, kept melanoma from coming back longer than the immunotherapy alone in a Phase 3 trial. The combination met both its primary endpoint on recurrence and a key secondary endpoint on whether the cancer spread to distant organs.

The topline, announced on the a16z podcast "Inside Moderna's Personalized Cancer Vaccine" by Moderna CEO Stéphane Bancel in conversation with a16z general partner Jorge Conde, is the first late-stage evidence that the mRNA modality behind Moderna's COVID vaccine can be turned into a per-patient cancer drug. The therapy, called intismeran autogene, is Moderna and Merck's individualized neoantigen therapy. Each dose is engineered from the patient's own tumor: clinicians sequence the tumor and healthy cells, identify the mutations most relevant to the cancer, then encode up to 34 of those mutations into a single mRNA written for that one patient.

"This is more than a decade of work culminating in an important milestone," Bancel said on the podcast, framed as both a company success and a test of the mRNA platform's reach beyond infectious disease.

That framing understates the manufacturing bet. Standard vaccines and cancer drugs are made in bulk lots. Moderna's oncology program makes one medicine per patient, on a clinical clock. Bancel said the company has compressed biopsy-to-treatment to roughly 42 days, a figure that doubles as a logistics claim and a competitive claim. The same per-patient mRNA design is being tested against lung, kidney, bladder, pancreatic, and gastric cancers, according to Bancel's LinkedIn post and the interview.

The trial, INTerpath-001, enrolled patients with melanoma whose tumors had been surgically removed. That is the highest-risk window for recurrence, which is why recurrence-free survival and distant metastasis-free survival were chosen as endpoints. Adding the personalized vaccine to Merck's pembrolizumab (marketed as Keytruda), the established immunotherapy, is the specific test that just read out positive. Recurrence-free survival means the cancer did not come back locally or regionally. Distant metastasis-free survival means it did not show up in organs like the lungs, liver, or brain.

Both endpoints matter, and both are why the readout is being treated as more than a press release. The company has not yet disclosed magnitudes: no hazard ratios, no subgroup breakdowns, no median follow-up time, no safety detail. Detailed clinical data is to be presented at an upcoming medical conference, Bancel said, and is not yet public.

That gap is the load-bearing caveat. Three things remain unresolved:

Durability. The trial met its endpoints, but how long the recurrence-free benefit holds, in months or years, is not in the topline.

Combo attribution. Keytruda alone is a powerful immunotherapy. Whether the personalized vaccine meaningfully added to it, or whether the combo's effect is closer to Keytruda's known performance, is the question every payer and oncologist will ask when the full data drops.

Generalizability. "Up to 34 neoantigens" and "roughly 42 days" are company figures. How often the manufacturing actually hits those marks across sites and patients will determine whether the per-patient model is a research triumph or a commercial product.

Moderna is also positioning the platform for rare genetic and autoimmune diseases, work the company has discussed for years. The topline matters because it converts that long arc from a story about a technology into a story about a treatment class that, if the full data holds, may reach oncology clinics in the next few years.

The next milestone is the medical conference presentation, where INTerpath-001's full results, including hazard ratios, subgroup data, and follow-up, will be public for the first time. That is when the durability and combo-attribution questions get answered, or get worse.